Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on November 9, 2026. According to the company's announcement, the two share the same underlying model. Fable 5.2 ships with stronger cybersecurity and biology safeguards and is open to all users. Mythos 5.2 has fewer safeguards and is limited to vetted organizations through Anthropic's Cyber Verification Program and Life Sciences Verification Program. The API model ID for Fable is claude-fable-5-2.
Anthropic says the new models are a step above Claude Opus 5.5 on nearly all of the capabilities it evaluated. Among its reported scores, Fable 5.2 reached 69.9% on FrontierSWE v2, against 62.3% for Opus 5.5 and 65.5% for GPT-6 Astra. It scored 49.1% on FrontierCode 2.0 at Xhigh effort, against 45.5% for Opus 5.5, and 74.2% on Humanity's Last Exam with tools. The model did not lead every benchmark: Gemini 4 Carbon scored higher on AutomationBench (53.9% to 51.4%), and Opus 5.5 scored slightly higher on AA-Briefcase. Anthropic also reports that at High effort, Fable 5.2 matches Opus 5.5's best FrontierCode score at roughly half the cost per task.
Anthropic lists standard pricing at $8 per million input tokens and $40 per million output tokens. A fast mode costs $16 and $80 respectively and generates tokens up to 2.5 times faster. The company says output is up to 30% faster than Fable 5.1 and Mythos 5.1. Fable 5.2 is available on Anthropic's platforms and through Amazon Web Services, Google Cloud and Microsoft Azure. Subscribers on eligible paid plans get extra Fable usage this week.
On security, Anthropic says Mythos 5.2 scored 89.6% on its internal CVE Rediscovery evaluation, against 78.3% for Mythos 5.1. Claude Enterprise customers can use Claude Security, which runs on Mythos 5.2. For Fable 5.2, Anthropic says its cyber and biology classifiers block about 55% fewer requests than they did for Fable 5.1. Blocked cyber queries are routed to Opus 4.8 and blocked biology queries to Opus 5.
The announcement says the risk of high-stakes misalignment is low. On an automated behavioral audit of about 1,900 scenarios, Mythos 5.2 scored 2.67 (lower is better), against 2.79 for Opus 5.5. Anthropic also notes that both models voiced suspicion that they were being evaluated more often than recent models, which it says complicates evaluation. External testers included the UK AI Security Institute and METR.
For developers, the practical changes are a new top-tier model ID, a fast mode on Claude Code, the Claude Platform, Google Cloud and Microsoft Foundry, and a lower refusal rate on legitimate security and biology work. Anthropic did not state a context window in the announcement.